A Smart IR Fire Detection System

Smart IR technology has advanced at a rapid pace over the last two decades, pushed along by primarily defense-related requirements. This includes such breakthroughs as Acousto-Optic Tunable Filters and microbolometer detector arrays. Coupled with the continuing growth in microprocessor power, this promises to make a new generation of fast, cheap and reliable infrared (IR) fire detectors possible that do not suffer from the false alarm problems that plague current designs. Since they sense the radiation emitted by a pending fire, they will be faster than smoke detectors that rely on smoke reaching the sensor. The proposed research is to see if the promise can be realized. Analysis will concentrate on the 4.3 um CO2 emission feature characteristic of flames. The 2.7 um H20 emission band will also be considered. Using computer models like the Aerodyne Radiation Code (ARC) supplemented with lab measurements on actual flames, the IR signature of flames and false alarm sources will be examined and compared. These will be used to drive a smart IR fire detection performance model that will explore the range of sensor designs for cost effective ones.